It’s Only Common Sense: How AI Consumes the Electronics Supply Chain
Dan’s Biz Bookshelf: ‘Building a Story Brand 2.0’
Aegis Software, a global provider of manufacturing execution, simulation and planning software, announced built-in Model Context Protocol (MCP) support in Simio. The new capability provides a standards-based way for preferred, compatible generative AI applications to work with Simio’s discrete-event simulation and advanced planning and scheduling software. Using natural language, teams can generate or refine model logic, understand how a Simio model works, troubleshoot errors, run and analyze experiments, and investigate production plans. This helps people who are new to Simio get started faster, enables experienced users to complete tasks more efficiently and gives teams a faster way to understand how operational decisions may perform before putting them into practice.
Simio’s existing AI capabilities include embedded neural networks that can learn from synthetic data generated during simulation runs and apply that learning to runtime decisions within the simulation model. The new built-in MCP support complements those capabilities by enabling organizations to connect preferred, compatible generative AI applications powered by large language models (LLMs) to Simio.
“AI is not new to Simio. Simio has been bringing AI and simulation together for years,” said Jason Spera, CEO of Aegis Software. “What the new built-in MCP support changes is how teams can work with Simio. AI-powered interaction can enable more people to contribute, help experienced users work faster and allow organizations to gain value from simulation and planning sooner.”
Connecting AI to How the Operation Actually Works
Large language models can answer a wide range of questions, but on their own they do not understand how a specific factory, hospital, airport, or distribution network will behave when demand shifts, a resource goes down or another condition changes. Simio captures the processes, resources, dependencies, constraints, and variability that influence how those operations perform.
Simio’s built-in MCP support provides a standardized way for compatible generative AI applications to work directly with Simio. Teams can use natural language to understand how a process works, investigate model logic, troubleshoot errors, and run or analyze experiments. For planning and scheduling, they can investigate plans and results, such as which orders finished late and what those orders have in common. The AI applications can also generate or refine model logic and draw on Simio documentation to answer questions about the software.
“AI-powered interaction can help people become productive with simulation sooner,” said Gregory Lang, Senior Director of Engineering and DevOps, Aegis Software. “Picture joining a team that has already built a simulation model of its operation. Understanding what someone else built takes time. With a compatible AI application connected through Simio’s built-in MCP support, you can quickly see how the model is structured, have its logic explained and ask questions about its behavior. Experienced users can use the same connection to troubleshoot issues and complete common tasks more efficiently.”
An Open Approach to AI
Simio’s built-in MCP support gives organizations flexibility in the AI technologies they choose. Rather than requiring a single AI provider, organizations can connect preferred, compatible generative AI applications through the Model Context Protocol.
Simio supports connection options including Claude Code, Codex, GitHub Copilot CLI and OpenCode, along with other MCP-compatible generative AI applications. Teams can also specify model instructions that provide additional context to the LLM when interacting with Simio. Instructions can include naming conventions, color rules and preferred terminology, helping align AI interactions with the structure and conventions of an individual model.
Built-in MCP support is available in Simio Release 20.294 or greater. Simio, part of the Aegis Software product family, will host a webinar on October 7, 2026, introducing the new capability and demonstrating how generative AI can support model development, troubleshooting, simulation experiments and production planning.
